Modifying Task Dependencies to Overlap Tasks
Fast-tracking
means starting the next task before its predecessor is complete. For
instance, you could tell your accountants to start counting the
chickens before all the eggs have hatched. When you identify partial
overlaps, you create them in Project by adding negative lag to the link between the tasks.
When the link lines in your project look like fishing line in a
tackle box, finding the right link line in the timescale is impossible,
so don’t even try. The Task Information dialog box is the easiest place
to add lag.
Here are the steps:
-
In the Gantt Chart table area, double-click the successor task.
The Task Information dialog box opens. -
Select the Predecessors tab.
All the predecessor tasks appear in the list, even those that aren’t on the critical path. -
Click the Lag cell for the predecessor you want to overlap, and then type a negative number for the overlap
.
For example, to overlap the two tasks by 2 days, type -2d. You can also overlap by a percentage of the predecessor task’s duration. For instance, if you type -25%
on a predecessor that takes 8 days, the overlap is 2 days. Keep in mind
that a percentage lag changes the length of the overlap when the
predecessor duration changes. -
Click OK.
The value appears in the Lag cell, and the task bars overlap in the timescale.
Paying More for Faster Delivery
Spending more money
to deliver in less time can make financial sense. A high-tech doodad
that will be obsolete in 2 years can’t afford a delay getting to
market. The sales you make could add up to more than the premium you
have to pay to finish the project earlier.
Crashing is the term project managers use for shortening
a schedule by throwing more resources at tasks. It’s not as dire as it
sounds, although it sometimes describes what happens when a horde of
people try to work on the same task. To the uninitiated, crashing
seems like an easy choice. If two people can build your website in 12
weeks, then four web developers should wrap it up in only 6 weeks, but
it doesn’t always work out that way.
It’s easy to assume that crashing won’t change cost—it looks like
you’re paying the same rate for the same number of hours. In reality,
there’s always a trade-off between time and money. When you crash
tasks, you must choose the tasks that shorten the schedule for the
least amount of money. That’s because every task has a magical duration
that results in the minimum cost. If the task runs longer, you spend
more money on things like office space, people’s salaries and benefits,
and keeping the freezer stocked with
Cherry Garcia ice cream. Ironically, shorter tasks can cost more,
too—for items like additional computers, higher-cost contractors, and
managing the larger team it takes to expedite the work.
Stakeholders might open the checkbook to crash a project, but they
don’t want to spend more than they have to. Shortening tasks that
aren’t on the critical path is a waste of money, because the project
duration doesn’t change a bit. At the same time, crashing doesn’t mean
shortening every task on
the critical path. Like cars, some tasks cost more to crash than
others. Your job is to shorten the schedule for the least amount of
money.
Crash tables help you compare the cost and time you can save by crashing
different tasks. For example, one task might cost an additional $50,000
to cut 2 weeks from the schedule, whereas another task cuts 2 weeks for
only $25,000. To choose cost-effective tasks, you must compare apples
to apples, that is, how much it costs for each week of duration you
eliminate. One method is to build a crash table in an Excel
spreadsheet. You can enter the duration you can cut and the cost, and
create an Excel formula that calculates the cost per week (or month or
day). The next section explains how to do that.
Long critical tasks are the best candidates for crashing,
since you can eliminate longer durations in fewer tasks. Shortening a
few long tasks may cost less—and save more time—than crashing a bunch
of shorter tasks.
Using a Spreadsheet to Choose Tasks to Crash
You could create custom fields in Project that hold crash costs and
durations, and calculate the crash cost per week with formulas you
define for the custom fields. However, it’s faster to move task data to
an Excel spreadsheet and calculate crash costs there.
Finding the
perfect combination of tasks to crash can get complicated. One task
might offer several weeks worth of low-cost crashing. However, if
shortening that task by 2 weeks takes it off the critical path, there’s
no point spending money to shorten it any more. The most effective
approach is to crash one task at a time (by adding resources to it in
Project), and then evaluate the critical path again to see what the
next step should be.
Since you start with the longest tasks and usually crash only a few,
start by sorting the critical tasks by duration, and then copy the
first several tasks from Project to Excel. Here are the steps for finding critical tasks to crash, and then copying them to Excel:
-
With a task-oriented view
displayed, click the down arrow to the right of the Task Name column
heading, and then choose Filters→Critical
.
To hide summary tasks, choose Format→Hide/Show, and then turn off the Summary Tasks checkbox. -
To list the longest critical
tasks first, click the down arrow to the right of the Duration column
heading, and then choose Sort Largest to Smallest.
The task IDs and task bars show up out of order, but the longest
tasks appear at the top of the task list. In addition to copying task
names to Excel, Duration and Cost come in handy if you want to sort
tasks in Excel by their original duration or cost.
Tip
The Entry table places the Task Name and Duration columns one after
the other. To simplify copying the fields you want, insert the Cost
field immediately after Duration. Right-click the Start column heading,
and then choose Insert Column. In the field name drop-down list, choose
Cost.
-
Drag from the Task Name cell of
the first critical task to the Cost cell for the last critical task you
want to copy, and then press Ctrl+C.
Project copies the cells to the Clipboard. -
In Excel, click the top-left cell where you want to copy the values, and then press Ctrl+V.
Excel pastes the values into worksheet cells starting at the
top-left cell and filling in cells to the right and down, as
illustrated in Figure 13. (If the values aren’t completely visible, widen the columns by dragging the dividers between the column headings.)
-
In the first three blank columns
to the right of the copied values, add headings for crash reduction,
crash cost, and crash cost per week
.
Fill in each row with the total amount of time you can shorten the task, and how much it costs to do so. -
In the first blank crash cost per week cell, type the formula to divide the crash duration by the crash cost (for instance, =
D2/E2).
You can then drag the copy handle over all the other crash cost per week cells so Excel can calculate the values. -
To find the least expensive tasks
to crash, sort the rows first by the amount of schedule reduction and
then by the crash cost per week.
In Excel, Choose Data→Sort. In the Sort dialog box, in the “Sort by”
drop-down list, choose Crash Reduction or the column name you entered.
In the Order drop-down list, choose Largest to Smallest. Click Add
Level to add a second sort criterion. In the “Then by” drop-down list,
choose Cost Per Week or the column name you entered. This time, in the
Order drop-down list, choose Smallest to Largest. Click OK. The tasks
appear with the longest reductions in duration first and the smallest
cost per week to the largest, as shown in Figure 13. -
Switch to Project and add resources to a task to shorten its duration.
Review the critical path to see whether the results are what you
want. For example, if the task is no longer on the critical path, you
may want to scale back on how much you crash it.
Repeat this step until the project duration is the length you
want—or the cost prohibits shortening it any further. At that point,
you may need to reduce the project scope.
|
